Why would you be confused about this? Temperature of a function of energy. We're talking about a global average temperature. Energy is being captured and retained. That raises the global average temperature. The specifics don't matter.
Predicting the weather on the other hand is very complicated because you're asking what specifically will happen in a specific place. The variables going into that are many and even if measured extremely accurately, it's not good enough because weather is chaotic.
This is reductionist, the specifics of what goes on at the surface do matter, that is the whole point of the climate change debate. How much chemical energy can we dig up and how do the byproducts affect the radiative ability of the planet (ie green house gases).
> How much chemical energy can we dig up and how do the byproducts affect the radiative ability of the planet (ie green house gases).
These are known stable parameters, we have a long recorded history of resource use and the insulative properties of gases in a mixture are tabled.
It's also well known in numerical modelling and physics why a number of systems have easy to predict long term coarse behaviour while also having short term impossible to predict fine grained behaviour, this exactly addresses your question about how can climate (coarse long term) be predicted when weather (short term, fine grained) is difficult.
See the Dzhanibekov Effect and the work of both Smale and Lorentz for insight.
A rotating (about intermediate axis) wing nut has a determined long arc trajectory of its CoG (centre of gravity) .. but an unpredictable short term tumble about its CoG.
